Accumulating evidence in animal and human studies indicates that pharmacotherapeutic responses are amenable to classical conditioning. It may be possible to create highly specific and effective placebo responses by means of initial exposure to a drug (acquisition) followed by placebo treatment interspersed with drug exposure at a sufficient frequency to prevent extinction of the conditioned response (maintenance). "Conditioned" placebos might thus be substituted for a proportion of active drug doses, an advantageous strategy for drug treatment that is limited by cumulative toxicity (e.g., the immunosuppressive or other cytotoxic treatments for AIDS, cancer, graft rejection and autoimmune disease). This project consists of three parts, each of which uses the treatment of hypertensive patients with atenolol as a model system for studying conditioned drug responses. The first part consists of a modified double crossover study in which a placebo-drug-placebo group is compared with a placebo-drug-nothing group. This part will verify the occurrence of a conditioned drug response (the difference in placebo response before and after drug exposure), and also refine a method for distinguishing conditioned drug effects from residual drug effects. The second part explores the feasibility and effectiveness of combining drug and placebo in chronic drug therapy. Decreasing daily doses of atenolol will be compared to stable drug doses administered with decreasing frequency, supplemented by placebo on the non-drug days. The latter strategy takes better advantage of conditioned drug effects, and should allow maintenance of a therapeutic effect at a lower cumulative drug dose than can the former strategy. This study will also yield useful observations regarding thresholds for drug doses and reinforcement schedules in maintaining therapeutic responses. The third part examines the acquisition and extinction of conditioned drug effects under continuous, partial, and "tapering" reinforcement schedules. All 3 parts will serve as prototypes in the development of a standard methodology for screening drug effects for their ability to be conditioned and for the mapping of conditioning parameters (i.e., characteristics of acquisition, maintenance and extinction). This foundational work is needed to make possible the practical application of this promising behavioral approach to pharmacotherapy.